heikal commited on
Commit
2c70928
·
verified ·
1 Parent(s): 5980afc

End of training

Browse files
README.md CHANGED
@@ -1,29 +1,16 @@
1
  ---
2
- base_model: openai/whisper-large-v3
3
- datasets:
4
- - heikal/arabic_call_splitted_6
5
  language:
6
  - ar
7
- library_name: transformers
8
  license: apache-2.0
9
- metrics:
10
- - wer
11
  tags:
12
  - generated_from_trainer
 
 
13
  model-index:
14
  - name: whisper large-v3 ar resturant 5
15
- results:
16
- - task:
17
- type: automatic-speech-recognition
18
- name: Automatic Speech Recognition
19
- dataset:
20
- name: Arabic calls splitted
21
- type: heikal/arabic_call_splitted_6
22
- args: 'config: ar, split: test'
23
- metrics:
24
- - type: wer
25
- value: 59.48275862068966
26
- name: Wer
27
  ---
28
 
29
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -32,9 +19,6 @@ should probably proofread and complete it, then remove this comment. -->
32
  # whisper large-v3 ar resturant 5
33
 
34
  This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the Arabic calls splitted dataset.
35
- It achieves the following results on the evaluation set:
36
- - Loss: 1.8319
37
- - Wer: 59.4828
38
 
39
  ## Model description
40
 
@@ -60,19 +44,16 @@ The following hyperparameters were used during training:
60
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
61
  - lr_scheduler_type: linear
62
  - lr_scheduler_warmup_steps: 500
63
- - training_steps: 1000
64
  - mixed_precision_training: Native AMP
65
 
66
  ### Training results
67
 
68
- | Training Loss | Epoch | Step | Validation Loss | Wer |
69
- |:-------------:|:-----:|:----:|:---------------:|:-------:|
70
- | 0.0005 | 250.0 | 1000 | 1.8319 | 59.4828 |
71
 
72
 
73
  ### Framework versions
74
 
75
- - Transformers 4.44.2
76
  - Pytorch 2.4.1+cu121
77
- - Datasets 3.0.0
78
- - Tokenizers 0.19.1
 
1
  ---
2
+ library_name: transformers
 
 
3
  language:
4
  - ar
 
5
  license: apache-2.0
6
+ base_model: openai/whisper-large-v3
 
7
  tags:
8
  - generated_from_trainer
9
+ datasets:
10
+ - heikal/arabic_call_splitted_7
11
  model-index:
12
  - name: whisper large-v3 ar resturant 5
13
+ results: []
 
 
 
 
 
 
 
 
 
 
 
14
  ---
15
 
16
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
19
  # whisper large-v3 ar resturant 5
20
 
21
  This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the Arabic calls splitted dataset.
 
 
 
22
 
23
  ## Model description
24
 
 
44
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
45
  - lr_scheduler_type: linear
46
  - lr_scheduler_warmup_steps: 500
47
+ - training_steps: 500
48
  - mixed_precision_training: Native AMP
49
 
50
  ### Training results
51
 
 
 
 
52
 
53
 
54
  ### Framework versions
55
 
56
+ - Transformers 4.45.1
57
  - Pytorch 2.4.1+cu121
58
+ - Datasets 3.0.1
59
+ - Tokenizers 0.20.0
config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "openai/whisper-large-v3",
3
  "activation_dropout": 0.0,
4
  "activation_function": "gelu",
5
  "apply_spec_augment": false,
@@ -7,10 +7,7 @@
7
  "WhisperForConditionalGeneration"
8
  ],
9
  "attention_dropout": 0.0,
10
- "begin_suppress_tokens": [
11
- 220,
12
- 50257
13
- ],
14
  "bos_token_id": 50257,
15
  "classifier_proj_size": 256,
16
  "d_model": 1280,
@@ -33,7 +30,7 @@
33
  "mask_time_length": 10,
34
  "mask_time_min_masks": 2,
35
  "mask_time_prob": 0.05,
36
- "max_length": 448,
37
  "max_source_positions": 1500,
38
  "max_target_positions": 448,
39
  "median_filter_width": 7,
@@ -43,7 +40,7 @@
43
  "pad_token_id": 50256,
44
  "scale_embedding": false,
45
  "torch_dtype": "float32",
46
- "transformers_version": "4.44.2",
47
  "use_cache": true,
48
  "use_weighted_layer_sum": false,
49
  "vocab_size": 51866
 
1
  {
2
+ "_name_or_path": "heikal/whisper-large-v3-ar-resturant5",
3
  "activation_dropout": 0.0,
4
  "activation_function": "gelu",
5
  "apply_spec_augment": false,
 
7
  "WhisperForConditionalGeneration"
8
  ],
9
  "attention_dropout": 0.0,
10
+ "begin_suppress_tokens": null,
 
 
 
11
  "bos_token_id": 50257,
12
  "classifier_proj_size": 256,
13
  "d_model": 1280,
 
30
  "mask_time_length": 10,
31
  "mask_time_min_masks": 2,
32
  "mask_time_prob": 0.05,
33
+ "max_length": null,
34
  "max_source_positions": 1500,
35
  "max_target_positions": 448,
36
  "median_filter_width": 7,
 
40
  "pad_token_id": 50256,
41
  "scale_embedding": false,
42
  "torch_dtype": "float32",
43
+ "transformers_version": "4.45.1",
44
  "use_cache": true,
45
  "use_weighted_layer_sum": false,
46
  "vocab_size": 51866
generation_config.json CHANGED
@@ -253,5 +253,5 @@
253
  "transcribe": 50360,
254
  "translate": 50359
255
  },
256
- "transformers_version": "4.44.2"
257
  }
 
253
  "transcribe": 50360,
254
  "translate": 50359
255
  },
256
+ "transformers_version": "4.45.1"
257
  }
model-00001-of-00002.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:948b4c8633e8861664d89236e785669a879002c60cac663c6be71defd62d0fa4
3
  size 4993448880
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0030b6806e7162549bfffb455281c83974f1a2986cf8a0ed7fb7d9eabbd1c092
3
  size 4993448880
model-00002-of-00002.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:21c2f60c5fa76df522249dc3c90ea62fad5fe5cc064544d042dc51aea14d18c6
3
  size 1180663192
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:29e0fb24544fe1baf160ef807a8d7daa39b120e4e591d4b0b0db722885f6aca5
3
  size 1180663192
runs/Sep29_15-52-46_ca0fdb65527a/events.out.tfevents.1727625176.ca0fdb65527a.3861.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c0ea6d073f142db102e0e15f5503faf98c71e46ec200a2b3e0ac76f0978a60e
3
+ size 10329
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:12f8b7ddad9b18aaa22d9e06db4508692a142b66b5d4b1222fce5b61af314ee0
3
  size 5432
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f4da889ec6ea5b8339f69761066415bdc2ab0107d1fd3f56aeb2fbb3d6e4d0b7
3
  size 5432